© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 117067
    Prev
    Next
    INDEX
    Explanations

    increases chances of survival

    np_acts-logits-general · gemini-2.5-flash-lite

    conditional statements showing possibilities, often using words like "will," "may," "can," or phrases indicating what something "helps" or "allows" you to do.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron strongly activates on modal and auxiliary expressions of possibility, ability, or recommendation (e.g. “can,” “may,” “should,” “will,” “allows,” “able,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    vidas
    -0.77
    nappropriate
    -0.75
    solving
    -0.74
     controla
    -0.73
     stopping
    -0.73
    `/
    -0.72
     fenomen
    -0.72
    vermectin
    -0.72
     Beni
    -0.71
    kJ
    -0.71
    POSITIVE LOGITS
     increase
    2.33
     increases
    2.25
     chances
    2.13
     zwięks
    2.03
    increase
    2.02
     increased
    2.00
     Increases
    1.92
     probability
    1.92
     odds
    1.90
     aumenta
    1.88
    Activations Density 0.066%

    No Known Activations